About
Zhiyue is a literature and research support tool for clinicians, graduate students and researchers. We combine the most time-consuming parts of literature work — searching, screening, quality tiering, methodological appraisal and first-draft writing — into one controllable, reviewable workflow, so you can reach a solid evidence base and a usable draft much faster.
What it does
- One-pass search across PubMed, Scopus and OpenAlex, with automatic de-duplication
- Journal tiering: filter by influence and disciplinary standing (Tier 1 / Tier 2) to locate high-quality evidence quickly
- Critical appraisal: methodological checklists for 11 study designs, including reviews, randomised controlled trials, cohort studies and meta-analyses
- Review drafting: quick and deep modes produce a structured first draft, with every citation traceable and verified after generation
- Export: Word / Markdown, plus formatted references
Quality and verifiability
- Citation-by-citation verification: after generation, every in-text citation is checked for existence (PMID/DOI) and semantic consistency, with the result reported at the end; citations that fail are flagged for human judgement.
- Programmatic reference rendering: the reference list is rendered deterministically from 7,501 EndNote journal templates rather than left to the model; export as RIS / BibTeX / EndNote XML / PMID list for direct import into a reference manager.
- Quality threshold: if fewer than 70% of citations pass verification, the generation is not charged.
- AI content labelling: both the interface and exported files (Word / Markdown) carry an AI-generated content label, and the label is written into file metadata as required by China's Measures for Labelling AI-Generated Synthetic Content; do not remove or obscure the label before sharing.
- Methodological appraisal: heuristic quality prompts based on 11 study-design frameworks (abstract-level, not a formal risk-of-bias assessment); a formal appraisal requires reading the full text and working through the tool's items by hand.
- Scope and limits: the output is an assistive first draft. The research question, inclusion/exclusion, methodological judgement and authorship remain entirely yours.
Background
The platform is built and maintained by clinicians working in research, and has been piloted in-hospital for literature searching, review writing and evidence-based discussion teaching with graduate students and residents.
Access and licensing
Register for trial credits; continued use is licensed by activation code, with credits metered per task type and itemised under Account & Credits.
Compliance and limits
- This platform isa literature-search and review-assistance tool, whose AI-generated content may contain errors and must be verified by the user. It does not replace the standard workflow of a systematic review or meta-analysis, and does not constitute medical advice or a basis for clinical decisions.
- The platformdoes not collect patient personal information, so please do not upload medical records or raw data containing patient identifiers.
- When using AI assistance in your writing, follow the AI-use and disclosure rules of your target journal and your institution.
Contact
Email: supports@zhiyue-ai.cn(feedback, partnership and licensing enquiries)